Commentary questions causal interpretation of ventricular unloading analysis during VA-ECMO

A commentary in Critical Care questions the causal interpretation of a multicenter target trial emulation comparing early left ventricular unloading with VA-ECMO alone in refractory cardiogenic shock. The underlying observational analysis assessed 60-day mortality using overlap weighting. The commentary reports no new data or analyses.

The cited weighted mortality risk difference was 1.4 percentage points (95% CI, −11.8 to 14.6), leaving clinically important benefit and harm compatible with the findings. The commentary highlights a possible mismatch between eligibility requiring at least 48 hours of ECMO support and follow-up starting at ECMO initiation. If applied, that criterion could exclude early deaths and early weaning. Assigning treatment according to unloading received within 24 hours also raises concerns because patients must survive until device placement to enter the unloading group.

Additional concerns include evolving hemodynamic indications for unloading that baseline weighting cannot address, and the near-confinement of intra-aortic balloon pump use to Paris and Impella use to Hamburg. These patterns complicate separation of device effects from center practices. The commentary calls for clarification of eligibility and reanalysis of the treatment grace period, interpreting the result as no detectable association in the weighted overlap population under local practices—not evidence of equivalence. These are methodological concerns rather than demonstrated corrections to the original findings; the commentary reports no funding or competing interests.